JP5811416B2 - Image processing apparatus, image processing method, and program - Google Patents

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program that execute predetermined image division processing.

  In general, when an area desired by a user (for example, a subject area) is extracted from an image, for example, when a part of a picture is extracted from an image of a hand-drawn picture on paper or the like, Otsu There is known a technique for partitioning an area by value conversion (see Non-Patent Document 1) or the like. However, there may be a case where the area cannot be partitioned appropriately due to variations in brightness of the entire image, uneven brightness, reflection, and the like. there were. In addition, as a technique for extracting an area, a technique called image segmentation includes various segmentation processing methods such as Grab Cut (see Non-Patent Document 2). However, it is assumed that an area to be extracted is designated. In addition, it is difficult to specify the region accurately, and the method of specifying the region also adversely affects the segmentation process.

Noriyuki Otsu's paper, Automatic threshold selection method based on discrimination and least squares criterion ", IEICE Transactions, Vol. J63-D, no. 4, pp. 349-356 (1980) C. Rother, V. Kolmogorov, and A. Blake, GrabCut: Interactiveforeground extraction using iterated graph cuts, ACM Trans.Graph., Vol. 23, pp. 309-314, 2004

  Therefore, conventionally, when a point in the extraction target area (subject) is tapped (specified by the user) on the image being displayed, it has a predetermined shape (for example, a circle) including the coordinate point corresponding to the tap position. An area is formed in the image, and the formed area is divided into background pixels and foreground pixels by performing segmentation processing using a program such as a graph cut (Grab Cut) as an initial foreground area. For example, a technique for performing a process of extracting a foreground portion is known (see Patent Document 1).

JP 2013-29930 A

  However, in the technique of the above-described patent document, it is possible to significantly reduce the burden on the user because the segmentation process can be executed by simply specifying one point of the extraction target. If the extraction target is a complicated shape, it may not be possible to fit the entire extraction target within an area of a predetermined shape (for example, a circle), which adversely affects the process of dividing the image into background and foreground pixels. There was a risk of affecting.

  An object of the present invention is to make it possible to accurately cut out a part of an area from an image without requiring an auxiliary operation for specifying the area by a user.

In order to solve the above-described problems, the image processing apparatus of the present invention provides:
Pixel value specifying means for specifying pixel values of some background pixels in the target image to be divided into foreground and background;
First specifying means for specifying pixels in the target image having pixel values corresponding to pixel values of the some background pixels as some background pixels;
Second specifying means for specifying other background pixels and foreground pixels in the target image using pixel values of each pixel in the target image and position information of the part of the background pixels;
As a result of specifying the background pixel and the foreground pixel by the second specifying means, if a background pixel area exists so as to be surrounded by the foreground pixel area, the area along the outermost contour in the foreground pixel area Cutting area setting means for setting as a cutting area in the target image;
It is provided with.

  According to the present invention, it is possible to accurately cut out a part of an area from an image without requiring an auxiliary operation for specifying the area by the user.

The block diagram which showed the basic component of the tablet terminal device applied as an image processing apparatus. (1) is a diagram illustrating image data (original image) read out as a processing target from the image memory 3c, and (2) is a diagram illustrating a state in which the original image is partitioned into a plurality of regions (16 regions). . The flowchart which showed the process for cutting out the picture out of the image which image | photographed the picture drawn by hand drawing on white paper. The flowchart following the operation | movement of FIG. (1), (2) is the figure which illustrated the luminance value histogram, respectively. (1), (2) is the figure which illustrated the luminance value histogram different from FIG. 5, respectively. The figure which showed the list which divided the luminance value into the group and counted the number of the luminance values which belong to each group. (1) is a diagram exemplifying an image image after executing a process of designating a luminance value near the peak of the luminance distribution as a background pixel, and (2) is a cut-out mask (cut-out) image obtained by the image dividing process. FIG. (1), (2) is a figure for demonstrating how to specify the outermost area | region (contour). (1) is a diagram illustrating a case where only the outermost region (contour) is adopted as a cutout mask image, and (2) is a diagram illustrating an image cut out by the cutout mask image.

Hereinafter, embodiments of the present invention will be described with reference to FIGS.
This embodiment illustrates a case where the present invention is applied to a tablet terminal device as an image processing device, and FIG. 1 is a block diagram showing basic components of the tablet terminal device.
The entire case of the tablet terminal device is, for example, a portable information terminal device of A5 size, and includes a touch input function, a wireless communication function, an Internet connection function, and the like. The control unit 1 which is the core of the tablet terminal device operates by supplying power from the power source unit (secondary battery) 2 and controls the overall operation of the tablet terminal device according to various programs in the storage unit 3. It has a central processing unit and memory.

  The storage unit 3 includes, for example, a ROM, a flash memory, etc., and a program memory 3a that stores a program for realizing the present embodiment in accordance with the operation procedure shown in FIGS. 3 and 4, and this tablet terminal It has a work memory 3b that temporarily stores various data and flags required by the apparatus, and an image memory 3c that stores various image data such as moving images taken by the camera. The storage unit 3 may include a removable portable memory (recording medium) such as an SD card or an IC card, and is connected to a network via a communication function (not shown). The state may include a storage area on a predetermined server device side. The image memory 3c is not limited to the built-in memory, and may be a removable portable memory such as an SD card. For example, the image memory 3c stores image data received and acquired via the Internet, or a camera device (not shown). For example, the photographed image data is stored.

  Although not shown in the figure, the operation unit 4 includes a power key for turning on / off the power, a selection key for selecting a moving image to be processed, and the like as push-button keys. A wireless LAN (Local Area Network) communication unit 5 is a wireless communication module capable of high-speed and large-capacity communication, and can be connected to the Internet or the like via a nearby wireless LAN router (not shown). The touch display unit 6 has a configuration in which a touch panel 6b is stacked on a display panel 6a, displays function names as software keys (soft keys), displays various icons, and displays image data. To do. The touch panel 6b of the touch display unit 6 constitutes a touch screen that detects a position touched with a finger or the like (including an operator such as a pen) and inputs coordinate data thereof. The method or the resistance film method is adopted, but other methods such as an optical sensor method may be used.

FIG. 2A is a diagram illustrating image data (original image) read out as a processing target from the image memory 3c.
The illustrated original image is, for example, an image (grayscale image) obtained by photographing a picture (demon face drawn with black lines) P drawn by hand using a black ink pen or the like in the approximate center of a white paper or whiteboard ), The brightness of the entire image varies due to reflections during shooting, uneven brightness, and the ceiling fluorescent light appears in the center of the right side of the image. This is an example of an image in which half is dark and the upper half is bright.

  In the figure, L1 and L2 indicate reflection portions of the fluorescent lamp, and are particularly bright. In the present embodiment, in order to cut out a picture P drawn on a white paper or a whiteboard from the original image, a portion having high luminance is specified as the background from the entire image. That is, in a case where an image obtained by photographing a picture (line drawing) P drawn in the second color (black) on a drawing medium having the first color (white) is used as the foreground. The portion close to the first color is specified as the background. In this case, before executing the image processing for cutting out the picture P from the original image, a luminance value histogram is generated in which the distribution of the luminance value (degree of brightness) of each pixel in the image is graphed. In the luminance distribution (brightness distribution), specify a luminance value near the peak (the luminance with the highest distribution) as the luminance of the background pixels instead of the highest luminance portion (the glossy reflection portion that is too bright). By executing a division process (for example, Grab Cut segmentation process), a desired picture P area (line drawing part area) is cut out from the image.

  This image dividing process is a process of dividing the entire image into foreground pixels and background pixels based on position information of some background pixels specified in the image, and is a segmentation process called Grab Cut. Note that the original image is not limited to a grayscale image, but may be a color image. In this case, luminance information may be generated from R (red), G (green), and B (blue) of the color image to obtain a luminance value histogram. In the present embodiment, a blank paper or a whiteboard is exemplified as the drawing medium. However, the drawing medium is not limited thereto, and a colored paper, a blackboard, or the like may be used.

FIG. 2B is a diagram showing a state in which the original image is equally divided into a plurality of regions, that is, 16 × 4 × 4 regions in a matrix.
In this embodiment, in order to deal with variations in brightness, uneven brightness, and the like in the entire original image, the entire original image is divided into 16 regions (1) to (16) equally. A luminance histogram is calculated and generated for each of the regions (1) to (16), and a segmentation process is executed by designating a luminance value near the peak of the luminance distribution as a background pixel. It should be noted that the number of sections and the way of partitioning are not limited to the case where the original image is divided into 16 equal parts, such as 8 equal parts and 32 equal parts.

  As described above, in this embodiment, the image processing apparatus (tablet terminal apparatus) processes the entire image based on the position information of some background pixels specified in the image (for example, an image obtained by photographing a picture drawn on paper). Division processing means (control unit 1, program memory 3a) for executing predetermined image division processing to divide into foreground pixels and background pixels, and identification means (control unit) for specifying the luminance of the background pixels based on the luminance distribution of the image 1. Program memory 3a) and control means (control for executing the predetermined image dividing process (Grab Cut segmentation process) by designating a pixel having luminance corresponding to the luminance specified by the specifying means as a background pixel Unit 1 and program memory 3a).

  Next, the operation concept of the image processing apparatus (tablet terminal apparatus) in the present embodiment will be described with reference to the flowcharts shown in FIGS. Here, each function described in these flowcharts is stored in the form of a readable program code, and operations according to the program code are sequentially executed. Further, it is possible to sequentially execute the operation according to the above-described program code transmitted via a transmission medium such as a network. In other words, in addition to the recording medium, an operation unique to the present embodiment can be executed using a program / data supplied externally via a transmission medium. 3 and 4 are flowcharts showing an outline of the operation of the characteristic part of the present embodiment in the overall operation of the image processing apparatus (tablet terminal device), and the flow of FIG. 3 and FIG. 4 is omitted. At that time, the process returns to the main flow (not shown) of the overall operation.

FIG. 3 and FIG. 4 are flowcharts showing a process for cutting out a part of a picture drawn by hand drawing on white paper. For example, the cut-out process is performed by a user operation from a menu screen. Execution is started when the image is selected, or when the clipped image is combined with another image, the execution is started as part of the image processing.
First, the control unit 1 takes in image data (for example, still image data) selected as a processing target from the image memory 3c as an original image (step S1 in FIG. 3). Then, the entire area of this image is divided into N equal parts, for example, equally divided into 16 areas as shown in FIG. 2 (2) (step S2), and a luminance value histogram for each divided area. Is calculated and generated (step S3). Then, the entire original image is designated as ROI (Region Of Interest) (step S4), and the luminance value near the peak of the luminance distribution is specified as the background pixel (step S5).

FIG. 4 is a flowchart for explaining in detail the process (step S5 in FIG. 3) for designating the luminance value near the peak of the luminance distribution as the background pixel.
First, the block number assigned to each area is reset to “0” in order to sequentially designate the areas (1) to (16) obtained by dividing the original image into 16 parts from the head (step S51). Then, it is checked whether the value of the block number is “16”, that is, whether or not the last 16th area (16) has been designated (step S52). Initially, the block number is “0” and the first Since the region (1) is designated, histogram data of the corresponding block (region) is acquired (step S53). Thereafter, the processes of steps S52 to S57 described later are repeatedly executed until the block number reaches “16”.

5 and 6 are diagrams exemplifying luminance value histograms.
As shown in FIG. 2 (2), among the regions (1) to (16) obtained by dividing the original image into 16 equal parts, FIG. 5 (1) shows a histogram of the region (1), and FIG. Indicates a histogram of region (7). FIG. 6 (1) shows a histogram of the region (10), and FIG. 6 (2) shows a histogram of the region (16). In the figure, the horizontal axis indicates the luminance value, the vertical axis indicates the number of pixels, and in the figure, the left side indicates dark (black) pixels, and the right side indicates bright (white) pixels.

  Here, as shown in FIG. 2, the region (1) is a region in the upper left corner, and is a region of only white paper or a whiteboard, and therefore, a luminance peak clearly appears. The region (7) is a region on the upper right side where the picture P is drawn. Since there are reflection portions L1 and L2 of the fluorescent lamp nearby, the luminance distribution is affected by the reflection. Has a higher brightness than its peak value. The area (10) is an area on the lower left side and is a part where the picture P is drawn. Since the area (10) is farther from the reflection parts L1 and L2 of the fluorescent lamp than the area (7), the luminance distribution is It is narrower than the region (7) and has a luminance lower than the peak value due to the influence of the black line inside the picture P. The region (16) is a region in the lower right corner and has a plurality of luminance peaks because it includes a large number of fluorescent lamp reflection portions L2.

  FIG. 7 shows a list in which the luminance values are divided into groups and the number of luminance values belonging to each group is counted. In the example shown in the figure, “group numbers” “0” to “0” are divided into 32 groups. In this configuration, “minimum range” and “maximum range” of “luminance value range” are associated with 31 ″. Now, taking the histogram of the region (10) shown in FIG. 6 (1) as an example, the following steps S53 to S57 will be described. First, the most distribution among the histograms of this region (10). A large luminance range is acquired (step S54). In this case, since the luminance peak in the histogram of FIG. 6A is the luminance value “16”, the luminance range of “group number” “16” is acquired as the luminance range with the largest distribution number.

  Then, the luminance group having the largest number of distributions and the group two higher than that are specified as the vicinity of the luminance peak (step S55). In this case, the group number “16” is specified as the luminance group having the largest distribution number, and the group numbers “17” and “18” are specified as the two higher groups as the vicinity of the luminance peak. The “brightness value range” of the group near the peak specified in this way is designated as the luminance of the background pixel (step S56). In this case, “brightness value ranges” “128” to “151” of the specified group numbers “16” to “18” are designated as the luminance of the background pixel. Then, after adding “1” to the block number and updating the value (step S57), the process returns to the above-described step S52, and the processes of steps S52 to S57 are repeated until the block number reaches “16”.

FIG. 8 (1) shows an image image after executing processing (step S5 in FIG. 3) for designating a luminance value near the peak of the luminance distribution as a background pixel for the original image shown in FIG. 2 (1). FIG.
This image shows the case where the background part is expressed in “black”. Other parts (white part and gray part) other than the black part are places where the background is not specified (parts where there is a possibility of line drawing) ).

  After specifying the luminance value of the background pixel for each region in this way, the process proceeds to step S6 in FIG. 3, and image division processing (Grab Cut segmentation processing) is performed for each region. In this image segmentation process, the first designated background pixel is a confirmed pixel that is confirmed to be the background, and for a target pixel whose foreground or background is undetermined, the luminance difference between adjacent pixels is determined in the image. This is a process for determining whether the foreground or the background is such that the total value obtained from a predetermined viewpoint for a plurality of portions becomes smaller. As a result of executing such image division processing, a cutout mask (cutting die) image as shown in FIG. 8B is obtained. In this case, the background pixel area exists so as to be surrounded by the foreground pixel area which is a picture (line drawing portion), and the inside of the picture drawn in the picture is also cut out. Only the region (outline) is adopted as a cut-out mask image. That is, an area along the outermost contour in the foreground pixel area is set as a cutout area in the image.

FIG. 9 is a diagram for explaining how to specify the outermost region (outline).
As a method of counting regions (contours), “openCv (trade name) cvFindContours” is used. That is, FIG. 9 (1) is an image in which the background is black and the white line drawing shows a quadrangle and a circle therein, and in the illustrated example, there are four black and white border lines (contour lines), that is, a quadrilateral shape. The case of the outer and inner boundary lines (contour lines) and the outer and inner boundary lines (contour lines) of a circle are shown. That is, the contour structure has a white contour on the outermost side, a black contour on the inner side, a white contour on the inner side, and a black contour on the inner side. FIG. 9 (2) illustrates how to count the contours. The first is the outer contour of the quadrangle (white contour), the second is the inner contour of the quadrangle (black contour), and the third is The outer outline of the circle (white outline), the fourth is the inner outline of the circle (black outline), and the first is the outer outline of the rectangle as the outermost area (contour) Therefore, only the outer contour line is left.

FIG. 10A is a diagram illustrating a case where only the outermost region (contour) is adopted as a cut-out mask image.
When the image cut-out process (step S8 in FIG. 3) is executed using the cut-out mask image created as described above, a cut-out image (image of a line drawing portion) as shown in FIG. 10 (2) is obtained. . The illustrated checkerboard pattern conceptually shows a transparent portion other than the cut-out image, and actually there is no checkerboard pattern. The cut-out image (line drawing portion image) obtained in this way is an image of only the picture P and includes an internal line drawing. Note that how to use the clipped image is arbitrary, but the clipped image may be combined in another image.

  As described above, in the present embodiment, the control unit 1 executes predetermined image division processing for dividing the entire image into foreground pixels and background pixels based on position information of some background pixels specified in the image. Prior to this, the luminance of the background pixel is specified based on the luminance distribution of the image, and the pixel having the luminance corresponding to this luminance is designated as the background pixel and the image division process is executed. Therefore, it is possible to accurately cut out a part of the area from the image without requiring an auxiliary operation. As a result, for example, an image obtained by shooting a hand-drawn picture using a black pen on a white paper or white board is affected by reflections during shooting, resulting in uneven brightness and uneven brightness. Even if this is the case, it is possible to accurately identify and accurately cut out the area of the picture.

  The control unit 1 divides the image into a plurality of regions, specifies the luminance of the background pixel for each region based on the luminance distribution for each of the divided regions, and has a luminance corresponding to the luminance specified for each region Is specified as the background pixel for each area and the image segmentation process is executed, so that even if there is a gradual difference in luminance overall, it is possible to accurately identify and properly cut out the picture area Become.

  Since the control unit 1 specifies the luminance value near the peak (the luminance with the highest distribution) as the luminance of the background pixel, not the highest luminance portion (the glossy reflection portion that is too bright) in the luminance value histogram. The image division process can be executed as a pixel whose foreground or background is undetermined without determining a glossy reflection portion that is too bright from the beginning as a background, and a picture area can be cut out more accurately.

  The original image to be processed is an image obtained by photographing a line drawing drawn in the second color on the drawing medium having the first color. When the line drawing portion is set as the foreground, the control unit 1 uses the first image. Since the portion close to the color is specified as the background, in the image obtained by photographing the picture (line drawing) drawn in the second color (black) on the drawing medium having the first color (white), this picture ( When the line drawing part) is used as the foreground, the part close to the first color can be used as the background, and the picture area can be cut out with higher accuracy.

  When the background pixel region is present so as to be surrounded by the foreground pixel region that is the line drawing part as a result of the image division processing, the control unit 1 follows the outermost contour in the foreground pixel region. Since the region is set as the cutout region in the image, it is possible to prevent the portion drawn in the picture from being cut out.

  In the image division processing, the first designated background pixel is a confirmed pixel that is confirmed to be the background, and for a target pixel whose foreground or background is undetermined, a difference in luminance between adjacent pixels is determined in the image. Since it is a process for determining whether the foreground or the background is such that the total value aggregated from a predetermined point of view for the portion becomes smaller, the foreground or the background is determined from the luminance difference between adjacent pixels with respect to the uncertain target pixel This makes it possible to cut out a picture area with higher accuracy.

  In the above-described embodiment, the case where the original image is a still image is illustrated, but the present invention can be similarly applied to a moving image. In this case, a region cutout process may be performed for each frame, or a region cutout process may be performed at predetermined frame intervals.

  In the above-described embodiment, the entire original image is equally divided into 16 areas. However, the number of sections is automatically determined based on the contents of the original image, or can be arbitrarily determined by a user operation. You may make it set.

  In the above-described embodiment, the area corresponding to the blank paper or the whiteboard in the image is used as the background, and the area corresponding to the picture is used as the foreground. However, the expression of the background and the foreground is convenient. For example, when the picture A area and the picture B area are divided, one picture may be regarded as a “background pixel” and the other picture as a “foreground pixel”. Further, the image to be cut out is not limited to a picture (line drawing part) but may be a subject.

  In the above-described embodiment, the segmentation method called Grab Cut is used as the predetermined image division processing, but other segmentation methods may be used.

  In the above-described embodiment, the case where the image processing apparatus is applied to a tablet terminal device has been described. However, the present invention is not limited to this, and a personal computer, a PDA (personal portable information communication device), a mobile phone such as a smartphone, and an electronic game It may be a music player or, of course, a digital camera itself.

  Further, the “apparatus” and “unit” shown in the above-described embodiments may be separated into a plurality of cases by function, and are not limited to a single case. In addition, each step described in the above-described flowchart is not limited to time-series processing, and a plurality of steps may be processed in parallel or separately.

The embodiment of the present invention has been described above. However, the present invention is not limited to this, and includes the invention described in the claims and the equivalent scope thereof.
Hereinafter, the invention described in the claims of the present application will be appended.
(Appendix)
(Claim 1)
The invention described in claim 1
Division processing means for performing predetermined image division processing for dividing the entire image into foreground pixels and background pixels based on position information of some background pixels specified in the image;
Specifying means for specifying the luminance of a background pixel based on the luminance distribution of the image;
Control means for designating pixels having brightness corresponding to the brightness specified by the specifying means as the partial background pixels and executing the predetermined image division processing;
An image processing apparatus comprising:
(Claim 2)
The invention according to claim 2 is the image processing apparatus according to claim 1,
Partitioning means for partitioning the image into a plurality of regions;
The specifying means specifies the luminance of a background pixel for each area based on the luminance distribution for each area divided by the dividing means,
The control means designates pixels having luminance corresponding to the luminance specified for each area by the specifying means as the partial background pixels for each area, and executes the predetermined image division processing.
An image processing apparatus characterized by this.
(Claim 3)
The invention according to claim 3 is the image processing apparatus according to claim 1 or 2,
The specifying unit specifies the luminance with the highest distribution as the luminance of the background pixel.
An image processing apparatus characterized by this.
(Claim 4)
The invention according to claim 4 is the image processing apparatus according to any one of claims 1 to 3,
The image is an image obtained by photographing a line drawing drawn in the second color on a drawing medium having the first color;
The specifying means specifies a portion close to the first color as a background when the line drawing portion is a foreground.
An image processing apparatus characterized by this.
(Claim 5)
The invention according to claim 5 is the image processing apparatus according to claim 4,
As a result of the execution of the predetermined image division processing, when a background pixel area exists so as to be surrounded by the foreground pixel area that is the line drawing portion, an area along the outermost contour in the foreground pixel area Is further provided with a clipping region setting means for setting as a clipping region in the image,
An image processing apparatus characterized by this.
(Claim 6)
According to a sixth aspect of the present invention, in the image processing apparatus according to the first to fifth aspects,
In the predetermined image division processing, the first designated background pixel is a confirmed pixel determined to be a background, and for a target pixel whose foreground or background is undetermined, the luminance difference between adjacent pixels is imaged. It is a process of determining whether the foreground or the background so that the aggregated value aggregated from a predetermined point of view for a plurality of parts is smaller.
An image processing apparatus characterized by this.
(Claim 7)
The invention described in claim 7
Identifying the luminance of the background pixels based on the luminance distribution of the image;
When performing predetermined image division processing for dividing the entire image into foreground pixels and background pixels based on position information of some background pixels specified in the image, the image has a luminance corresponding to the specified luminance. Designating a pixel as the partial background pixel and executing the predetermined image segmentation process;
The image processing method characterized by including.
(Claim 8)
The invention according to claim 8 provides:
A program for controlling a computer of an image processing apparatus,
The computer,
Division processing means for performing predetermined image division processing for dividing the entire image into foreground pixels and background pixels based on position information of some background pixels specified in the image;
Specifying means for specifying the luminance of a background pixel based on the luminance distribution of the image;
Control means for designating pixels having luminance corresponding to the luminance specified by the specifying means as the partial background pixels and executing the predetermined image division processing;
It is a computer readable program that is made to function as.

DESCRIPTION OF SYMBOLS 1 Control part 3 Memory | storage part 3a Program memory 3c Image memory

Claims (11)

Pixel value specifying means for specifying pixel values of some background pixels in the target image to be divided into foreground and background;
First specifying means for specifying pixels in the target image having pixel values corresponding to pixel values of the some background pixels as some background pixels;
Second specifying means for specifying other background pixels and foreground pixels in the target image using pixel values of each pixel in the target image and position information of the part of the background pixels;
As a result of specifying the background pixel and the foreground pixel by the second specifying means, if a background pixel area exists so as to be surrounded by the foreground pixel area, the area along the outermost contour in the foreground pixel area Cutting area setting means for setting as a cutting area in the target image;
An image processing apparatus comprising: The pixel value specifying means specifies pixel values of some background pixels based on a histogram of pixel values of a target image to be divided into foreground and background.
The image processing apparatus according to claim 1. The target image is an image obtained by photographing a line drawing drawn in the second color on a drawing medium having the first color;
The pixel value specifying unit specifies a pixel value corresponding to the first color as a pixel value of a background pixel when the line drawing portion is a foreground.
The image processing apparatus according to claim 1 , wherein the image processing apparatus is an image processing apparatus. The clipping region setting means determines that when a background pixel area exists so as to be surrounded by a foreground pixel area that is the line drawing portion as a result of specifying the background pixel and the foreground pixel by the second specifying means. A region along the outermost contour in the pixel region is set as a cutout region in the target image ;
The image processing apparatus according to claim 3 . The second specifying unit estimates a background pixel and a foreground pixel in the target image with a predetermined accuracy using a pixel value of each pixel in the target image and position information of the part of the background pixels. 5. The image processing apparatus according to claim 1 , wherein a process of specifying a background pixel and a foreground pixel in the target image is repeatedly executed with higher accuracy using the estimation result. Partitioning means for partitioning the target image into a plurality of regions;
The pixel value specifying means specifies pixel values of some background pixels for each area based on a histogram of pixel values for each area partitioned by the partitioning means,
The first specifying means specifies some background pixels for each region based on pixel values of some background pixels specified for each region;
The said 2nd specific | specification part specifies the background pixel and the foreground pixel of the said whole target image using the positional information on the one part background pixel specified for every said area | region . The image processing apparatus according to any one of the above. The pixel value specifying means specifies a pixel value having the most distribution as a pixel value of a background pixel;
The image processing apparatus according to claim 1 , wherein the image processing apparatus is an image processing apparatus. The pixel value specifying means specifies the luminance of some background pixels based on a luminance histogram of a target image to be divided into foreground and background,
Said first specifying means, in any one of claims 1 to 7, wherein the identifying the pixels in the target image having a luminance corresponding to the luminance of the background pixels in the part as part of the background pixel The image processing apparatus described. The second specifying means sets the first designated background pixel as a confirmed pixel determined to be the background, and for a target pixel for which the foreground or the background is undetermined, the difference in luminance between adjacent pixels is included in the image. To determine whether the foreground or background so that the aggregated value aggregated from a given point of view for multiple parts of the
The image processing apparatus according to claim 1 , wherein the image processing apparatus is an image processing apparatus. A process of identifying pixel values of some background pixels based on a histogram of pixel values of a target image to be divided into foreground and background;
A process of identifying pixels in the target image having pixel values corresponding to pixel values of the some background pixels as some background pixels;
A process of specifying other background pixels and foreground pixels in the target image using pixel values of each pixel in the target image and position information of the part of the background pixels;
As a result of specifying the background pixel and the foreground pixel, when a background pixel area exists so as to be surrounded by the foreground pixel area, an area along the outermost contour in the foreground pixel area is defined in the target image. A process to set as a crop area,
An image processing method comprising: The computer of the image processing device
Pixel value specifying means for specifying pixel values of some background pixels based on a histogram of pixel values of a target image to be divided into foreground and background;
First specifying means for specifying pixels in the target image having pixel values corresponding to pixel values of the some background pixels as some background pixels;
Second specifying means for specifying other background pixels and foreground pixels in the target image using pixel values of each pixel in the target image and position information of the part of the background pixels;
As a result of specifying the background pixel and the foreground pixel by the second specifying means, if a background pixel area exists so as to be surrounded by the foreground pixel area, the area along the outermost contour in the foreground pixel area Clip region setting means for setting the clip region as a clip region in the target image,
Program to function as.

Images